Critical

Sociology
http://crs.sagepub.com/

What Does the Ruling Class Do When It Rules? Some Reflections

On Different Approaches To the Study of Power in Society
Gran Therborn
Crit Sociol 1999 25: 224
DOI: 10.1177/08969205990250021101
The online version of this article can be found at:
http://crs.sagepub.com/content/25/2-3/224.citation

Published by:
http://www.sagepublications.com

Additional services and information for Critical Sociology can be found at:
Email Alerts: http://crs.sagepub.com/cgi/alerts
Subscriptions: http://crs.sagepub.com/subscriptions
Reprints: http://www.sagepub.com/journalsReprints.nav
Permissions: http://www.sagepub.com/journalsPermissions.nav

>> Version of Record - Mar 1, 1999

What is This?

Downloaded from crs.sagepub.com at UNIV MASSACHUSETTS AMHERST on September 15, 2012

WHAT DOES THE RULING CLASS DO WHEN IT RULES?

SOME REFLECTIONS ON DIFFERENT APPROACHES TO
THE STUDY OF POWER IN SOCIETY
Gran Therborn
What is the place of power in society? What is the relationship
between class and power? Answers diVer, as is to be expected, given
the obvious signi cance of class and power to the evaluation of a given
society. The question itself, however, appears simple and straightforward enough. Ideological biases apart, what seems to be at issue is the
famous question of scienti c method, of what is the most adequate
method to answer the question.1 But is the question really so clear and
simple? From what we know about paradigms (Kuhn) and problematics (Althusser) of science is it very likely that, for example, a proletarian revolutionary and critic of political economy (Marx), a German
academic historian and sociological follower of Austrian marginalism
(Weber), a descendant of JeVersonian democracy (Mills), an admirer of
contemporary liberal economics (Buchanan-Tullock, Parsons), or an
adherent of some of the ruling political ideas of present-day USA (Dahl,
Giddens[2]), would be concerned with the same problem and ask the
same questioneven when they use the same words?
Leaving subtler points and distinctions aside we can distinguish at
least three diVerent major approaches to the study of power in society.
The rst and most common one we might call the subjectivist approach.
With Robert Dahl it asks: Who governs?,3 or with William DomhoV:
Who rules America?,4 or in the words of a British theorist of strati cation, W.G. Runciman: who rules and who is ruled?,5 or in the militant pluralist variant of Nelson Polsby: Does anyone at all run this
community? 6
This is a subjectivist approach to the problem of power in society
not in the same sense as subjective in the so-called subjective conceptions of strati cation, which refer to strati cation in terms of subjective evaluation and esteem, in contrast to strati cation in terms of,
say, income or education. It is a subjectivist approach in the sense that
it is looking for the subject of power. It is looking, above all, for an answer
to the question, Who has power? A few, many, a uni ed class of families, an institutional elite of top decision-makers, competing groups,
Critical Sociology 25,2/3
Downloaded from crs.sagepub.com at UNIV MASSACHUSETTS AMHERST on September 15, 2012

what does the ruling class do?

225

everyone, or no one really? The focus of the subjectivists is on the

power-holding and power-exercising subject.7
The common subjectivist question can then be studied and answered
in various ways. This has in fact given rise to a very lively methodological as well as substantial debate in the United States in the fties
and sixties, which still has not been superseded, between the pluralists and the elite and the ruling class theorists. 8 Essentially, it has been
a debate within the framework of liberal political ideology and liberal
political theory, accepting the liberal conception of democracy as a starting-point and then investigating whether the contemporary manifestations of liberal democracy, in the present-day United States or in other
Western countries, correspond or not to that conception. But it has also
included important contributions from Marxist authors, who have basically con ned themselves within this framework, accepting battle on the
terrain chosen by the enemy.9 The latter case, by the way, highlights
the far-reaching eVects of prevailing ideology, shaping even the form of
opposition to itself.
Outside the subjectivist fold and its internal polemics about diVerent
methods and answers, another type of question is raised by some authors
who base themselves on liberal economic ideology and liberal economic
theory. We might label it the economic approach. In the businessmans
manner, the question here is not who, but how much. Power is regarded
above all as a capacity to get things done. The primary emphasis is on
power to rather than power over and the crucial question is not
the distribution but the accumulation of power. As a theory of power
the economic approach features two main variants, a sociological and
a utilitarian. The main proponent of the former is Talcott Parsons.
Parsons conceives power as a circulating medium, analogous to money10
and de nes it as generalized capacity to secure the performance of
binding obligations by units in a system of collective organization when
the obligations are legitimized with reference to their bearing on collective goals and where in case of recalcitrance there is a presumption
of enforcement by negative situational sanctionswhatever the actual
agency of that enforcement. 11
In the utilitarian economic theories of democracy, little attention
and consideration is allotted the phenomenon of power and its conceptualization. Politics is seen from the perspective of an individualist
theory of collective choice and the meaning of power is then derived
from the assumed blessings of market exchange. This approach, write
Buchanan and Tullock, incorporates political activity as a particular
form of exchange; and, as in the market relation, mutual gains to all parDownloaded from crs.sagepub.com at UNIV MASSACHUSETTS AMHERST on September 15, 2012

226

gran therborn

ties are ideally expected to result from the collective relation. In a very
real sense, therefore, political action is viewed essentially as a means
through which the power of all the participants may be increased, if we
de ne power as the ability to command things that are desired by men.12
Although they can be said to share a common approach to power,
inspired by liberal economics, concentrating as they do on non-con ictual
power to, the two main variants of the economic approach also show
diVerences that are by no means insigni cant. In the sociological variant, power is generated and operates in social relationships, whereas in
the utilitarian conception it is basically a non-relational asset. In both
the problem of class and power by and large disappears.
With little of the elaborate theoretical imagination of the above-mentioned authors, the economic approach to power has also been applied
to the problems of political development and modernization, above
all by Samuel P. Huntington. Huntington starkly emphasizes the importance of the accumulation of power over the question of its distribution. He opens his book Political Order in Changing Societies by proclaiming,
The most important political distinction among countries concerns not
their form of government but their degree of government. The diVerences
between democracy and dictatorship are less than the diVerences between
those countries whose politics embodies consensus, community, legitimacy, organization, eVectiveness, stability, and those countries whose
politics is de cient in these qualities.13 Modern political systems diVer
in the amount of power in the system, not in its distribution. 14 To
Huntington it is the general liberal ideas about economic development,
rather than liberal economic theory, which provides the model.
The third approach might be named a structural-processual approach.
But with its focus on society as an objective structured totality and on
contradiction, motion, and change, we had perhaps better call it the
dialectical-materialist approach, embodied in the new scienti c study of history and society founded by Marx, historical materialism. Here the primary focus is on the historical social contexts and modalities of power,
and the rst question is: What kind of society is it? Then: What are the
eVects of the state upon this society, upon its reproduction and change?
The central task of Capital was not to identify those who have the
wealth and those who are poor, nor those who rule and those who are
ruled, but, as the author pointed out in his preface, to lay bare the
economic law of motion of modern society. That is, Marx was above
all interested in how wealth and poverty, domination and subjugation
are being (re)produced and how this can be changed. The basic focus
of study is on neither property nor the property owners but on capital,
Downloaded from crs.sagepub.com at UNIV MASSACHUSETTS AMHERST on September 15, 2012

what does the ruling class do?

227

that is, on (particular historical) relations of production and their relationship to the productive forces and to the state and the system of ideas.
II
This third approach to the problem of power in society owes its more
roundabout character to the fact that it seriously and systematically tries
to tackle two fundamental problems largely neglected by the other
approaches. One concerns power to, the other relates to power over.
One question which should be seriously faced is: Power to do what?
What is a particular amount of power used for? The utilitarian answer
to maximize ones utilityis hardly very satisfactory in view of the enormous variety of historical social forms, and thereby systems, of power.
For the same reason we do not learn very much from Parsons discussion of power in terms of realization of collective goals.15 Nor should
it be assumed a priori, or made a part of the de nition, that, as Parsons
contends, power is exercised in the interest of the eVectiveness of the
collective operation as a whole,16 rather than in the interest of the
exploitation of one class by another.
What power to means depends on the kind of society in which it
operates. A Marxist analysis of a given society rst of all focuses on its
mode(s) of production, its system(s) of relations and forces of production.
By determining the relations of production the Marxist analyst at the
same time determines if there are classes in the given society and what
classes there are, because classes in the Marxist sense are people who
occupy certain positions in society as basically de ned by the relations
of production. If immediate productionin husbandry, agriculture, industry, transport, etc.and the appropriation and control of the surplus
produced are separated among diVerent role incumbents, and are not
united in an individual or in a collective, there are classes. And the
diVerent modes of separation (slavery, feudalism, capitalism, etc.) mean
diVerent classes.17
Determining the relations of production does not pertain only to the
context of political power. It is also directly related to the question of
power, since the separation between the immediate producers and the
appropriators of the surplus product entails speci c relations of domination and subordination.18 Exploitative relations of production directly
involve relations of domination, and in what may be called the key passage of Marxs materialist interpretation of history he says, The speci c
economic form in which unpaid surplus labour is pumped out of the
direct producers, determines the relation of domination and servitude,
Downloaded from crs.sagepub.com at UNIV MASSACHUSETTS AMHERST on September 15, 2012

228

gran therborn

as it emerges directly out of production itself and in turn reacts upon

production. Marx then continues and makes his basic proposition about
the relationship between the economy and the polity (the meaning, truth
and fruitfulness of which proposition is under debate): Upon this basis,
however, is founded the entire structure of the economic community,
which grows up out of the conditions of production itself, and consequently its speci c political form. It is always the direct relation between
the masters of the conditions of production and the direct producers
which reveals the innermost secret, the hidden foundation of the entire
social edi ce and therefore also of the political form of the relation
between sovereignty and dependence, in short of the particular form of
the state.19
For the adherents of the subjectivist approach, in both its pluralist
and elitist variants, to raise the problem of power to do what? means
to ask: What do rulers do when they rule? Where do the leaders lead the led? To
say or imply that what rulers do when they rule is to maintain their
ruling position is at best trivialand not infrequently wrong. Intentionally
and unintentionally what rulers do and do not do aVects the ruled, and
the same sort of power subjectsin terms of personal background and
present interpersonal relationsmay aVect the ruled in very diVerent
ways. There are diVerent eVects under pluralism or elitism, diVerent
eVects under, say, military governments and centralized oligarchical
organizations. 20 And there are many ways for a ruling class to exercise
and maintain its rule, other than by supplying, from its own ranks, the
political personnel. It may therefore be argued that rulers and ruling
classes would be better identi ed not by their names and numbers, their
social background and power career although all this is of course not
without importancebut by their actions, that is by the objective eVects
of their actions. From this perspective, the Marxist interjects into the
subjectivist discussion, polarized around democracy and dictatorship or,
in its contemporary, somewhat lower-pitched versions, pluralism and
elitism: democracy of what class, dictatorship of what class?
There is a second aspect to what rulers do when they rule. Talcott
Parsons once made a famous critique of utilitarianism for its inability
to account for social order.21 What all kinds of subjectivist elite and ruling class theorists are unable to do is to account for social change.
Characteristically enough, the classical elite theorists, who really thought
out the consequences of their theories, all basically held that society did
not change. Instead they drew a picture of an eternal cycle of rising,
ruling, degenerating, and falling elites. This goes for all of them,
Gumplowicz, Mosca, Pareto, and Michels.22 Ultimately they tended to
Downloaded from crs.sagepub.com at UNIV MASSACHUSETTS AMHERST on September 15, 2012

what does the ruling class do?

229

reduce people and human society to biology.23 Now, though men certainly are biological organisms, it is an obvious fact that human society has changed over the ages of its existence and has taken a number
of forms. The task of a social science must necessarily be to analyse
these diVerent historical forms and their change. This cannot be done
by taking the subjects of power, their psyche, their will, as the starting
point, but only by taking the social context in which they rule.
The society in which the rulers rule contains certain possibilities and
tendencies of change. The rulers rule in a certain stage of development
of a certain social structure, and their rule both aVects and is aVected
by the tendencies and contradictions inherent therein. The subjectivists
stop before analysing these tendencies and contradictions and typically
conclude: Look, only a few have power, that is bad! Or: Look many
have power, that is good! It should be noted that the important thing
in this context of power and change is the eVect of the rulenot directly
upon individuals, nor upon the gains and handicaps it means for persons and groups24but upon the social structure and social relationships in which the individuals live, because it is the latter, rather than
the sheer fact of being handicapped or exploited, which determine the
possibilities of change and revolt.
Besides the problem of power to, there is also a very important
neglected problem involved in power over. Are the diVerent moments
of the exercise of somebodys power over somebody else related to each
other? If we assume neither that social life is completely random and
unpatterned, nor that it is a uni ed, consensual, collective operation,
then how should the relationships be studied and how can they be
grasped?
At rst sight it might appear gratuitous to call this a neglected problem, as it is precisely what the substantial polemics between pluralists
and elitists have been about. True to their common subjectivist core,
however, the pluralists and the elitists have concentrated on a secondary
aspect of the problem. What they have been debating is whether there
is an interpersonal relation between the diVerent moments of power in
society: Is there a cohesive elite which unites the diVerent exercises of
power by making the decisions in diVerent areas? Or is there a fragmentation of decision-making between diVerent little or not-connected
groups? What this formulation of the issue does not take into serious
account is that an interpersonal fragmentation of decision-making does
not necessarily mean that the diVerent power events are random and
unpatterned. On the contrary, it is a basic, and it seems, warranted,
assumption of social science that the events in human society are in
Downloaded from crs.sagepub.com at UNIV MASSACHUSETTS AMHERST on September 15, 2012

230

gran therborn

some ways always patterned, and therefore possible to grasp by scienti c

analysis. What the elitism-pluralism theorists have been doing, then, is
to concentrate on the existence or non-existence of one possible form
of the patterning of power in society and, it should be added, on a
form which hardly seems to be the most important one in modern complex societies.
Little is gained in answering this kind of objection by referring to
another kind of interpersonal identity than that ensured by overlapping
membership in cohesive power groupsby referring, that is, to a common identity of ideas, to a consensus of values.25 How is a particular
kind of consensus and its maintenance to be explained,26 and how does
it actually operate, so general and abstract as it tends to be in modern
societies? What objective social structures and social relationships are
brought about and/or maintained, how are peoples lives patterned by
the diVerent exercises of supposedly consensual power?
Important methodological critiques of pluralism have been developed
by Bachrach and Baratz,27 and most recently by Lukes,28 with their
inclusion of institutional mobilization of bias and of non-decisionmaking,29 and in the case of Lukes, of latent con icts and of the eVects
of inaction. 30 But they do not deal with the present problem, of power
over. In fact, the subjectivist orientation of these authors seems to preclude a way out for the elitists in this respect. What their re ned methods can do is to detect more hidden manifestations of elite rule, but
they can hardly nd social patternings of exercise of power other than
those of a uni ed power subject. With Bachrach-Baratz this is strongly
implied by their conception of power, and its related concepts, as an
interpersonal relation between A and B.31 With Lukes it follows from
the authors moralistic concern with responsibility. For this reason Lukes
is uninterested in impersonal forms of domination and wants to concentrate on cases where it can be assumed that the exerciser(s) of power
could have acted diVerently from how they did. And in this context he
throws in a distinction between power and fate!32 To Lukes too, then,
power should be analysed primarily with a view to nding subjects of
power, identi able, free, and responsible originators of acts (and nonacts). He seems to remain stuck within the pluralist-elitist framework,
of either a uni ed elite or various elites or leadership groups (whose
interrelationship as a relation of power over others remains obscure,
unless they themselves are aware of their relationship).
Marx opened up a path out of the pluralist-elitist impasse, one which
seems to have remained almost completely unnoticed among sociologists
and political scientists, including writers who have explicitly referred to
Downloaded from crs.sagepub.com at UNIV MASSACHUSETTS AMHERST on September 15, 2012

what does the ruling class do?

231

Marx, more or less critically. The radical novelty and dissimilarity to

others of the Marxian approach seems to have been drowned in subjectivist receptions and reinterpretations. The way out indicated by Marx
is that the study of a given society should be not just a study of its subjects nor of its structure only, but also and at the same time should be
an inquiry into its process of reproduction. Signi cantly, it is in the study
of the process of reproduction that Marx analyses the class relationships
of exploitation and domination.
Capitalist production, therefore, under its aspect of a continuous connected process, of a process of reproduction, produces not only commodities, not only surplus-value, but it also produces and reproduces
the capital relation; on the one side the capitalist, on the other the
wage-labourer. 33 In a critique of the subjectivist conceptions of market
exchange in 18th- and 19th-century economics Marx provided a critique in advance of 20th-century sociologists as well: To be sure, the
matter looks quite diVerent if we consider capitalist production in the
uninterrupted ow of its renewal, and if, in place of the individual capitalist and the individual worker, we view them in their totality, the
capitalist class and the working-class confronting each other. But in so
doing we should be applying standards entirely foreign to commodity
production. 34
For the study of power in society the perspective of reproduction
means that the commanding question of all the variants of the subjectivist approachWho rules, a uni ed elite or competing leadership
groups? Is the economic elite identical with or in control of the political elite?is displaced by the question: What kind of society, what fundamental relations of production, are being reproduced? By what
mechanisms? What role do the structure and actions and nonactions of
the state (or of local government) play in this process of reproduction,
furthering it, merely allowing it, or opposing it?
The analysis of reproduction makes possible an answer to the question of how the diVerent moments of the exercise of power in society
are interrelated, even if there is no conscious, interpersonal interrelation. They are interrelated by their reproductive eVects. A given kind
of relations of production may be reproduced without the exploiting
(dominant) class de ned by them being in control of the government
in any usual and reasonable sense of the word, even though the interventions of the state further and/or allow these relations of production
to be reproduced. And yet the fact that a speci c form of exploitation
and domination is being reproduced, is an example of class rule and is
an important aspect of power in society.
Downloaded from crs.sagepub.com at UNIV MASSACHUSETTS AMHERST on September 15, 2012

232

gran therborn
III

The limited aim of this paper is to distinguish between diVerent

approaches to the problem of class and power, particularly between the
dialectical-materialist (Marxist) approach and the variants of the subjectivist approach. Such a distinction seems important in order to open
up possibilities for the application of the speci c Marxist approach, given
the fatal aws of the prevailing subjectivist one. The distinction is particularly important at the present juncture in the social sciences, where,
in spite of a renewed interest in and acknowledgement of Marx, an
evaluation of the truth and fertility of Marxist theory tends to be made
impossible by the amalgams currently fashionable in post-1968 sociology. In such eclectic constructionswhich appear to be made according to a recipe like, one part Marx, two parts Weber, and two parts
more recent sociology (including ingredients supplied by the cook himself ), seasoned with diVering amounts of hot (radical) and mild (liberal)
spicesthe distinctly Marxist analysis is drowned.
With such an aim, the present paper is not a direct contribution to
the study of class and power. But within this limitation I will nally
try to indicate a few guiding threads for a Marxist type of empirical
investigation of the problem of class and power. That only rather general and tentative guiding threads will be oVered re ects, I think, not
only the limitations of the present paper and of its author, but also the
fact that Marx opened up a radically new scienti c path, to be constantly cleared of the lush vegetation of dominant ideologies, and on
which only the very rst steps in the direction of systematic theory have
been taken.
The primary object of empirical study, for a grasp of the relations
of class, state, and power, should be neither interpersonal relations
between diVerent elites (for instance, the government and the business
elites), nor their social backgrounds, nor issues and decisions and nondecisionsalthough all this is important. The primary object should be
the eVects of the state on the (re)production of a given (whether found
or hypothesized) mode (or modes) of production. The relations of domination entailed by the relations of production are concentrated in the
state. Through the state the rule of the ruling class is exercised. The
character of this rule has to be grasped from the eVects of the state.
There are two aspects to these eVects: what is done (and not done) through
the state, and how things are done that are done through the state. We
need a typology of state interventions and a typology of state structures.

Downloaded from crs.sagepub.com at UNIV MASSACHUSETTS AMHERST on September 15, 2012

what does the ruling class do?

233

The typology of state structures should distinguish among the diVerential

eVects (of legislative, administrative, and judiciary arrangements and procedures, of mechanisms of governmental designation, of organization of
army and police, etc.) upon the extent to which the state can be used
by diVerent classesthat is, their eVects on whether and to what extent
the rule of a given class of people (with certain characteristics and
quali cations as de ned by their position in society) can operate through
the state structure under investigation. 35 In this way broad types of state
structures can be identi ed and distinguished in terms of their class
character, for example feudal states, bourgeois states and proletarian
states (in which the principle of politics in command, as realized in
soviets, workers parties, mass movements of cultural revolution, etc.,
seems to be a central characteristic). Various speci c state apparatuses,
such as legislative bodies, the judiciary, or the army, could also be studied from this point of view. It should of course not be assumed that a
concrete state at a speci c point in time necessarily has a homogeneous
class character in all its institutionswhich raises the problem then of
how to establish its dominant class character.
To study the process by which the state actually operates we also
have to have a typology of state interventions (including non-interventions signi cant to the (re)production of given relations of production).
Such a typology could be almost endlessly re ned. Basically, however,
it should comprise two dimensions. One concerns what is done, and
the other how it is done. In other words, one refers to the external
eVects of state intervention on other structures of society, above all on
the relations of production, (but also on the ideological system), and the
other refers to the internal eVects upon the state itself. State intervention can either further, merely allow, or go against, and at the limit
break, given relations of production.
And they can either increase, maintain, or go against, and at the
limit break, given relations of political domination as embodied in the
character of apparatuses of administration (and government) and repression. (The possibilities of successfully breaking given relations of production are fundamentally determined by the particular stage of the
relations and forces of production, and the stage of the relations of force
between classes which this implies.) The following table illustrates the
types of state intervention possible along these two dimensions.

Downloaded from crs.sagepub.com at UNIV MASSACHUSETTS AMHERST on September 15, 2012

234

gran therborn
EVect upon
given
relations of
political
domination
(Structure of
administration and
repression)
Increase
Maintain
Go against/
Break

EVect upon given

relations of production

Further

Allow

Go against/
Break

1
4
7

2
5
8

3
6
9

This typology can be applied both to a given political measure, such

as a social security program, nationalization, a land reform, a school
reform, etc., and to the sum of actions undertaken by a given government over a given period. It is in this way that the class character, in
the Marxist sense, of a regime or a policy is to be ascertained. For example, a nationalization act or a land reform can allow and even further capitalist relations of production if it is carried out through the
rules of the capitalist game, involving compensation more or less at market value, implementation through the established legislative and administrative procedure, and the creation of enterprises run by new owners
using wage labor for pro t (or for subsidizing other enterprises run for
pro t). But such measures can also be put into eVect in the opposite
manner, without necessarily meaning the complete abolition of capitalism in the society. A regional policy can be carried out with the help
of various kinds of subsidies, such as tax rebates, to capitalist enterprises, thus following the logic of capitalist relations of production but
making a certain localization of plants more pro table. But the same
measure can also go against that logic, through mandatory planning.
The class character is determined on the basis of the identity of the
dominant (exploiting) class (i.e., the dominant class of the particular relations of production furthered or allowed by the interventions). If there
is a discrepancy in the eVects upon the relations of production and the
structure of the state, this indicates a contradictory and unstable situation. For instance, in the case of the last period of Czarist Russia, the
state furthered the developing capitalist relations of production while at
the same time basically maintaining a largely pre-capitalist form of state;
Soviet Russia in the 1920s allowed capitalist relations of production to
develop while maintaining a proletarian dictatorship; and the Allende
Downloaded from crs.sagepub.com at UNIV MASSACHUSETTS AMHERST on September 15, 2012

what does the ruling class do?

235

regime in Chile partly allowed, partly went against, capitalist relations

of production while maintaining the existing state structure (its administration, judiciary, and army).
It should be underlined that, as a rule, there are a number of ways
in which given relations of production can be furthered in a given situation. Opinions therefore usually diVer over which is the best one.
Consequently, a given state intervention may very well go against the
current opinion of business organizations, but still further capitalist relations of production. The bourgeoisie as a class, and its interests, are not
identical with the identity or ideas of a particular group of business
leaders at a particular point in time. From this perspective we can
understand the pattern that frequently appears in capitalist politics,
wherein policies, when rst introduced, are opposed by business groups
and conservative parties, but once carried out, are accepted by them,
with longer or shorter delay (e.g. collective bargaining, social security
programs, Keynesian economics). This phenomenon is hidden by the
issues-and-decisions approach of the pluralist methodologists.
The ruling class of a given society is the exploiting class of that exploitative system of relations of production furthered (above others, if there
are other relations of production in the society) by the content and form
of the totality of state interventions during a given period. The ruling
class need not necessarily be the economically dominant class, in the
sense of the exploiting class of the dominant mode of production in a
society where there are several modes of production (e.g., self-subsistence farming, feudalism, petty commodity production, capitalism).
One possible re nement of the typology is to distinguish among their
eVects on the two diVerent classes (exploiting and exploited) that bear
the exploitative relations of production that the interventions in question further, allow, or go against. For instance reformist governments
usually are to be found in squares 4 and 5 in the table above although
certain of their measures will be found in 1 and 2, as for example antistrike measuresbut a more re ned typology would direct the study to
their possible eVects on the relations of distribution within the given relations
of production. Another re nement, as regards the eVects on the state,
would be to diVerentiate between the class eVects on the administrative
and on the repressive apparatuses of the state. Fascist regimes, analyzed
in terms of their eVect on capitalist relations of production, belong in
square 1, but they are more closely characterized within that type by
their increase in the repressive apparatus of the bourgeois state.36 A third
elaboration would be to distinguish between eVects of the state on different fractions of capital, e.g., industrial versus banking (or commercial or agrarian) capital, domestic versus foreign capital, big (monopoly)
Downloaded from crs.sagepub.com at UNIV MASSACHUSETTS AMHERST on September 15, 2012

236

gran therborn

versus small capital. In this way diVerent hegemonic fractions of the

bourgeoisie can be identi ed.
What the ruling class does when it rules, in the Marxist sense, then,
is not to make, as a compact unit, all important decisions in society.
The rule of the ruling class is exercised by a set of objectively interrelated but not necessarily interpersonally uni ed mechanisms of reproduction, through which the given mode of exploitation is reproduced.
The ruling class, in this sense, is not a uni ed power subject. The rule
of the ruling class is not necessarily, and is usually not, expressed in
conscious collective decisions and actions by the class as a whole. What
the ruling class does when it rules is not primarily a matter of subjective intentions and actions. Its rule is embodied in an objective social
process, through which a certain mode of production is maintained and
expanded, guaranteed and furthered by the state. This means that the
pluralist-elitist debate does not pertain to the existence of a ruling class
in the radically diVerent Marxist sense. What that debate is concerned
with are certain aspects of the mode of organization of the ruling class,
such as its cohesion.
It should be noticed that neither the existence of a ruling class, nor
what class is the ruling class, nor the amount of its power, are de ned
here a priori. What classes there are has to be uncovered by an analysis of the relations of production in a given society. The ruling class
has to be identi ed and the amount of its political power, the range of
its rule, has to be ascertained by a study of the structure and the interventions of the state. The dialectical-materialist approach to power in
society is an empirical approach, although of a quite diVerent kind.
Having located the ruling class, another task is then to lay bare the
mechanisms of its rule, which includes nding an answer to why the
actual interventions of the state functionas such mechanisms.
The state power of the ruling class is part of the total reproduction
process of society. As Poulantzas37 has pointed out, there are two aspects
of reproduction (and it should be added of revolution as well): the reproduction of the positions of the given social structure, and the reproduction of individuals who can occupy them. For example, capital, wage
labor, and capitalist enterprise have to be reproduced, as does the state
apparatus. The reproduction of position also involves, at least in the
long run, the production and reproduction of a compatibility between
the diVerent levels of the social structure. The reproduction of capitalism requires not only the reproduction of capitalist enterprise but in the
long run the reproduction of a compatible capitalist state as well.

Downloaded from crs.sagepub.com at UNIV MASSACHUSETTS AMHERST on September 15, 2012

what does the ruling class do?

237

But also, new generations of individualsand the given individuals

year in and year outhave to be trained to occupy the given positions,
to be quali ed or subjected to ful ll adequately the tasks provided by
the social structure. Out of the new-born infants a given proportion
have to be reared to become owners and managers of capital, other
portions to become workers, white collar employees, and administrative
and repressive personnel, or petty-bourgeois farmers, shop-keepers, and
artisans.
What broad types of mechanisms of reproductionwithin which we
can seek and nd the concrete mechanisms in concrete societiescan
be identi ed? One of primary importance is, of course, economic constraint. Economic constraint functions, in ways laid bare by speci c economic analysis, in and through the stage of the productive forces, the
inherent dynamics of the relations of production, and the interdependence of the forces and relations of production. It operates on various
levels and decisively aVects both the reproduction of positions and of
the agents to occupy them. A given level of the development of the
productive forces excludes certain relations of production, makes them
untenable or obsolete and non-competitive; and the necessity for some
kind of material reproduction then favors certain other relations of production, and determines the range of political options, such as for the
Bolshevik government after the civil war. On a lower level, economic
constraint imposes certain limits upon what a capitalist corporation or
a feudal manor can do to stay in business, limits for instance on the
extent to which one corporation or manor can tamper with the capitalist and feudal relations of production governing other corporations
or manors. Economic constraint operates in a constant process to reproduce a certain structure of economic positions, by sanctions of bankruptcies, unemployment, poverty, and sometimes outright starvation.
Economic constraint is an important mechanism for keeping even revolutionarily-conscious peasants and workers in line and harnessing them
for the reproduction of the society they would like to overthrow.
Another important type of mechanism of reproduction is political,
and includes two basic subtypes, administration and repression, which
in modern societies are both regularly concentrated in a distinct state
apparatus (or, rather, system of state apparatuses). Through administrative interventionstaxation, regulations, subsidies, countercyclical policies, etc.the reproduction of a certain mode (or modes) of production
is favored or hindered. Administration also functions in the reproduction of agents for the positions of the given modes of production, through

Downloaded from crs.sagepub.com at UNIV MASSACHUSETTS AMHERST on September 15, 2012

238

gran therborn

such things as manpower policies (from binding peasants to their landlords to stimulating labor market mobility) and social security policies
(from providing dreaded workhouses to supplying social security bene ts,
which function both to alleviate dangerous discontent and to stimulate
business). Administrative interventions operate to ensure the overall compatibility of the substructures of society. The mechanism of administration also includes mechanisms for the reproduction of the state apparatus
itself, embodied in constitutional provisions, procedures for the due handling of issues, or legal conceptions. These can hinder a government
which may intend on far-reaching social change, or can restrict the
accessibility to the state of certain classes or sections of classes.
Repression is the other important political mechanism of reproduction. The development or maintenance of certain modes of production
can be repressed by the army, the police, prisons, or the executioner.
Movements of opposition can be repressed in various ways and degrees.
(One interesting and neglected object of study in this respect concerns
to what extent the development of the labor movement in the United
States has been stopped by repression, especially after World War I and
World War II.) Individuals who refuse to accept any of the given positions
can be taken care of, for instance in prisons or in mental hospitals.
Mechanisms of reproduction, then, are not only, nor even mainly,
ideological, as sociologists are prone to assume.38 But ideological mechanisms are of course important too. Their primary role is not in legitimating the prevailing system,39 but rather in a diVerential shaping of
aspirations and self-con dence and in a diVerential provision of skills
and knowledge. This process of quali cation and subjection, in which
little human animals are formed into members of diVerent classes, takes
place in a number of ideological apparatuses: the family, the educational system, the church, the mass media, on-the-job training, and the
workplace (where so much of the inculcation of hierarchy and discipline takes place).40 These apparatuses and the dominant ideological formation which takes place in them, are not necessarily congruent with
each other. One particularly problematic relationship is that between
the family and other apparatuses, such as the church and (above all in
modern capitalism) the educational system. On the one hand, the family is an important mechanism of reproduction; but on the other, a certain amount of individual mobility is crucial to the reproduction of the
system. For individual mobility implies that the commanding positions
are occupied by more competent persons, as well as oVering an obvious channel of discontent. As Marx pointed out, referring both to capitalist enterprise and to the Catholic Church of the Middle Ages: The
Downloaded from crs.sagepub.com at UNIV MASSACHUSETTS AMHERST on September 15, 2012

what does the ruling class do?

239

more a ruling class is able to assimilate the most prominent men of the
dominated classes the more stable and dangerous is its rule.41 The
Marxist perspective notes that rigidly diVerentiated access to the educational system tends to make exploitation less stable. In the Marxist
perspective, what is most important to the reproduction of exploitation
is not diVerential access to the educational system, but the diVerential
educational system itself. Mobility, then, is essentially an ideological
mechanism of reproduction. So also is another phenomenon dear to all
subjectivists, interpersonal intercourse, which contributes to a common
outlook among the representatives of diVerent constituencies.
Through these mechanisms of reproduction the ruling class can exercise its rule and keep state power without necessarily having to supply
the political and administrative personnel. The economic laws of motion
of a given society set a very high threshold for their possible trespass
by politicians. The structural arrangements of the state (its class character) circumscribe the state interventions decided upon by the government.
The ideological mechanisms of reproduction shape both the politicians
even labor politicians with no personal intercourse with the bourgeois
creamand the population at large, including the exploited classes.
All these mechanisms operate in and through the con ict and struggle of classes. Class struggle then does not mean, even mainly, battles
between uni ed, self-conscious entities. It means con ict and struggles
between people who occupy diVerent positions in exploitative modes of
production.
Reproduction and revolution, consequently, are not to be understood
in terms of mechanisms of reproduction versus class struggle. The reproductive mechanisms also produce, at the same time, mechanisms of revolution. To realize this is, of course, a basic feature of a dialectical
approach. Marx analyzed, for instance, how the expanded reproduction
of capital also meant the development of contradictions between the
relations and the forces of production. That analysis might be extended
to the political and ideological processes of reproduction. For example,
the strengthening of the stateand with it the strengthening of administrative and repressive operations of the statewhich characterizes the
modern, imperialistic state of capitalism, has been accompanied by more
devastating contradictions among capitalist states. The two world wars
of the 20th century gave rise to non-capitalist regimes among a third of
humanity. Similarly, at the ideological level, the role of the intelligentsia,
both in old Russia and China and recently in the advanced capitalist
societies, testi es to the fact that the mechanism of quali cation and subjection might also take on the character of a revolutionary mechanism,
Downloaded from crs.sagepub.com at UNIV MASSACHUSETTS AMHERST on September 15, 2012

240

gran therborn

developing a contradiction between quali cation and subjection. There

are also mechanisms of revolution which operate in and through the
class struggle. And, looked at from the other side of the same coin, the
class struggle is fought in and through mechanisms of reproduction and
revolution. But all that is another part of the story, and, maybe, part
of another paper.
Notes
The original version of this article appeared in The Insurgent Sociologist 6:3 (Spring,
1976), pp. 316.
1. See, e.g., R. Dahl, A Critique of the Ruling Elite Model, American Political Science
Review (APSR) 52 (1958), pp. 46369; N. Polsby, How to Study Community Power:
The Pluralist Alternative, Journal of Politics 22 (1960), pp. 47484: P. Bachrach
M. Baratz, The Two Faces of Power, APSR 56 (1962), pp. 94752; P. Bachrach
M. Baratz, Decisions and Nondecisions: An Analytical Framework, APSR 57 (1963),
pp. 64151; R. Merelman, On the Neo-Elitist Critique of Community Power, APSR
62 (1968), pp. 45160; APSR 65 (1971), pp. 106380; F. Frey, Comment: On Issues
and Nonissues in the Study of Power, APSR 65 (1971), pp. 10811101; R. Wol nger,
Rejoinder to Freys Comment, APSR 65 (1971), pp. 110204. An overview can be
gained from the reader edited by R. Bell. D. Edwards, and H. Wagner, Political Power
(New York: Free Press, 1969).
2. According to Giddens the USA is the most democratic advanced society in the
world: A. Giddens, The Class Structure of the Advanced Societies (London: Hutchinson, 1973),
p. 175. For a comment on the issue, see below. Dahls conception of the prevailing
regime in USA is expounded in, among other places, R. Dahl, Pluralist Democracy in the
United States: Conict and Consent (Chicago: Rand McNally, 1967).
3. R. Dahl, Who Governs? Democracy and Power in an American City (New Haven: Yale
Univ. Press, 1961).
4. W. DomhoV, Who Rules America? (Englewood CliVs, N.J.: Prentice-Hall, 1967).
5. W.G. Runciman, Towards a Theory of Social Strati cation, in F. Parkin, ed.,
The Social Analysis of Class Sructure (London: Tavistock, 1974), p. 58.
6. Polsby, op. cit., p. 476.
7. In the egalitarian orientation of Bachrach-Baratz this focus is coupled with a lookout for who, if any, gain and who, it any, are handicapped by the existing mobilization of bias. Besides their above-mentioned articles see their Power and Poverty (New
York; Oxford University Press, 1970).
8. The substantial debate includes, from the elitist side, F. Hunter, Community Power
Structure (Chapel Hill: Univ. of North Carolina Press, 1953): C.W. Mills, The Power Elite
(New York: Oxford Univ. Press, 1956); W. DomhoV, op. cit., Bachrach-Baratz, op. cit.:
M. Parenti, Power and Pluralism: A View From the Bottom, in M. Surkin A. Wolfe,
eds., An End to Political Science (New York: Basic Books, 1970), pp. 11143; M. Creson,
The Un-Politics of Air Pollution (Baltimore: The John Hopkins Press, 1971). Among the
contributions of the pluralists are S.D. Riesman, et al., The Lonely Crowd (New York:
Doubleday Anchor, 1953); R. Dahl, Who Governs?; E. Ban eld, Political Inuence (New York:
Free Press, 1961). For references to the methodological discussion see above note 1.
9. The most important example is R. Miliband, The State in Capitalist Society (London:
Weidenfeld & Nicolson, 1969). The polar opposite kind of Marxist stance is exempli ed
by N. Poulantzas, Political Power and Social Class (London: NLB, 1973). In a well-argued
article the latter has been criticized for not coming to grips with, and thus not really
revealing the weaknesses of, the problematic of his opponents: E. Laclau, The Speci city
Downloaded from crs.sagepub.com at UNIV MASSACHUSETTS AMHERST on September 15, 2012

what does the ruling class do?

241

of the Political: The Poulantzas-Miliband Debate, Economy and Society (1975), pp. 87110.
Although mainly restricted to a distinction between diVerent approaches to the problem
of power, the present article tries to take account of the criticisms of both Poulantzas
and Miliband. At the same time I am indebted to them both for their very valuable
contributions.
10. T. Parsons, On the Concept of Political Power, in idem, Sociological Theory and
Modern Society (New York: Free Press, 1967), p. 306.
11. Ibid., p. 308.
12. J. Buchanan G. Tullock, The Calculus of Consent (Ann Arbor: Univ. of Michigan
Press, 1962), p. 23. Anthony Downs somewhat less sanguine view of power does not
refer to power over either, but to unequal power to, because of inequalities of information and income. A. Downs, An Economic Theory of Democracy (New York: Harper &
Row, 1957), pp. 25758.
13. S. Huntington, Political Order in Changing Societies (New Haven: Yale Univ. Press,
1968), p. l.
14. Ibid., p. 144.
15. Parsons, op. cit., p. 308.
16. Ibid., p. 318.
17. I have developed an analysis of the Marxist concept of class in my Klasser och
okonomiska system (StaVanstorp: Cavefors, 1971). Cf. the rst part of my Classes in Sweden
1930 70, in R. Scase, ed., Readings in the Swedish Class Structure (London: Hutchinson,
forthcoming).
18. Power in a society should of course be studied not only in terms of the nonspeci c, extra-organizational power of organizational elites, but also in terms of the mode
of organization itself, particularly the mode of organization of peoples working lives,
which diVer both in the kind and the amount of domination and independence. However,
the Marxist focus on exploitation and class is related to the discussion of power only in
the broad sense of the latter, in the sense of A signi cantly aVecting B in a situation of
possible negative sanctions against Bs non-compliance. The speci cation of power in
terms of responsibility, choice, and agreement, and distinctions between fate, coercion,
authority, manipulation, and power, are internal to a subjectivist discourse and as such
are outside the Marxist analysis proper. The latter does not start from the point of
view of the actor but from ongoing social processes.
19. K. Marx, Das Kapital (Otto Meissner, 1921), III:2, p. 324; T. Bottomore
M. Rubel, eds., Karl Marx. Selected Writings in Sociology & Social Philosophy (London: Watts
& Co., 1956), p. 99.
20. Consider, for instance, the sterility of the Michels type of organizational theory
when faced with the completely diVerent behavior of Social Democratic and Communist
parties in August 1914 and September 1939, respectively.
21. T. Parsons, The Structure of Social Action (New York: McGraw-Hill, 1937).
22. See G. Therborn, Class, Science and Society (Goteborg: Revopress, 1974), ch. 4.2;
English edition, London: NLB, 1976 forthcoming (American distributor: Humanities
Press).
23. Pareto extended the theory of class struggle into the thesis that The struggle
for life and welfare is a general phenomenon for living beings. V. Pareto, Les Systemes
Socialistes (Paris, 190203), 11, p. 455. Michels referred to the struggle between organized workers and strikebreakers in terms of struggle for feeding-ground. R. Michels,
Political Parties (Glencoe: Free Press, 1958), p. 307.
24. This is in contrast to the approach of Bachrach-Baratz, op. cit.
25. Dahl has written, . . . democratic politics is merely the chaV. It is the surface
manifestation, representing super cial con icts. Prior to politics, beneath it, enveloping
it, restricting it, conditioning it, is the underlying consensus . . . among a predominant
portion of the politically active members. R. Dahl, A Preface to Democratic Theory (Chicago:
Univ. of Chicago Press, 1956), p. 132. But what if consensus is the surface manifestation of something else, enveloping, restricting and conditioning electoral politics?
Downloaded from crs.sagepub.com at UNIV MASSACHUSETTS AMHERST on September 15, 2012

242

gran therborn

26. This is a weak spot in the otherwise well-substantiated critique of pluralist these
by Miliband (op. cit.). Miliband basically shrinks from really analysing governments whose
personnel is not recruited from the economic elite, and where the higher echelons of
the administration may also be recruited otherwise. In such cases he merely refers to
the ideology of the political leaders as part of a bourgeois consensus (see ch. 4, part IV).
He does provide some empirical material and suggestions for a study of the problem,
but it is fundamentally outside his model of control. For the analysis of advanced bourgeois democracies, of reformism fascism, and military governments, a more complex
model seems crucial. Similarly, the important works of William DomhoV, on the hautebourgeois backgrounds and connections of American politicians and administrators, and
on the cohesiveness of the top-most stratum of the US bourgeoisie, would bene t from
being located in a much more elaborate conceptualization and analysis of the US power
structure and of the contradictory development of US society.
27. Bachrach-Baratz, op. cit. (1962, 1963, 1970).
28. S. Lukes, Power: A Radical View (London: Macmillan, 1974).
29. A nondecision means a decision that results in suppression or thwarting of a
latent or manifest challenge to the values or interests of the decision-maker, BachrachBaratz, op. cit. (1970), p. 44.
30. Lukes, op. cit., chs. 4, 7. Lukes draws upon the work of Crenson, op. cit.
31. Bachrach-Baratz, op. cit. (1970), ch. 2.
32. Lukes, op. cit., pp. 5556. Cf. Marx: I paint the capitalist and the landlord in
no sense coleur de rose. But here individuals are dealt with only in so far as they are the
personi cations of economic categories, embodiments of particular class-relations and
class interests. My standpoint, from which the evolution of the economic formation is
viewed as a process of natural history, can less than any other make the individual
responsible for relations whose creature he socially remains, however much he may subjectively raise himself above them. Das Kapital, I, p. viii; Capital (London: Lawrence &
Wishart, 1970), vol. 1, p. 10. Marxs view certainly did not mean that the power of the
capitalist was a fate to submit to, but something that could be combatted and abolished.
It does mean, however, that it is rather pointless to accuse the capitalists of not behaving like non-capitalists. The Marxian standpoint implies, of course, that the arm of criticism is replaced by the criticism of arms (i.e., the class struggle in all its forms).
33. Marx, op. cit., I. p. 541; Lawrence & Wishart editions, I. p. 578.
34. Ibid., p. 549 and p. 586, respectively.
35. This seems to indicate a way out of the dilemma posed by Claus OVe in his very
penetrating essay, Klassenherrschaft und politisches System. Zur Selektivitat politischer
Institutionen, in his Strukturprobleme des kapitalistischen Staates (Frankfurt: Suhrkamp, 1972).
This is an objectivist approach to the problem of the selectivity of the state, but it is
not based on any de nitions of the objective interests of the revolutionary class, which
OVe rejects (p. 86). Neither does it mean that an empirical inquiry into the class character of the state only can be made post festum, as OVe concludes (p. 90), when the class
struggle has developed to the point where the limits of a given state appear.
36. Fascism is also distinguished by its furthering of monopoly capitalist relations of
production, which points to still another distinction in terms of fractions of classes furthered or disadvantaged by the state interventions. Cf. N. Poulantzas, Fascism and Dictatorship
(London: NLB, 1974).
37. N. Poulantzas, On Social Classes, New Left Review 78 (1973), p. 49.
38. Parsons treats the problem of reproduction or pattern-maintenance solely in
terms of transmission of values. For a relatively recent formulation see T. Parsons. The
System of Modern Societies (Englewood CliVs, N.J.: Prentice-Hall, 1971), pp. 1015. Similarly
all social sources of stability singled out by Parkin (op. cit., 1971, ch. 2) refer to ideological mechanisms: mobility, the educational system, religion, gambling, and the fostering of beliefs in luck. A noteworthy exception is H.F. Moorhouses interesting account
of the political and economic constraints imposed upon the British working class up to

Downloaded from crs.sagepub.com at UNIV MASSACHUSETTS AMHERST on September 15, 2012

what does the ruling class do?

243

1918 and their role in shaping later working class deference, in his The Political
Incorporation of the British Working Class: An Interpretation, Sociology 7 (1973), pp.
34159. See also the discussion by R. Gray, The Political Incorporation of the Working
Class, Sociology 9 (1975), pp. 10104; H.F. Moorhouse, On the Political Incorporation
of the Working Class: Reply to Gray. Ibid., pp. 10510. See footnote 72, on the concentration (in the discussion of social reproduction) on legitimation, a preoccupation coming out of the Weberian tradition.
39. To identify the ideological mechanisms of reproduction with the processes of legitimation would imply that people do not revolt against the given rule under which they
live because they regard it as legitimate. This seems hardly warranted. People may not
revolt, political and economic constraints aside, because they do not know the kind of
domination they are subjected to. That is, they may be hold ignorant not only of its
negative features but of its positive claims and achievements as well. They may be ignorant of alternatives, or they may feel themselves incapable of doing anything about it,
even if they know of other possible types of societies. But this ignorance, disinterest, and
lack of con dence are not simply there, as characteristics of certain individuals and
groups, they are produced by de nite social processes. See the important distinction
between pragmatic and normative acceptance made by Michael Mann, The Social
Cohesion of Liberal Democracy. American Sociological Review 35 (1970), pp. 42339. The
one-eyed concentration on legitimation is often related to a normative conception: that
every rule should be based on the true and knowing consensus of the ruled, thereby
holding it legitimate. See, for instance, J. Habermas, Legitimationsprobleme im Spatkapitallismus
(Frankfurt: Suhrkapm, 1973), esp. pp. 162V. But that is another question. Interestingly
enough, Habermas and OVe both accept Webers ideal type of competitive capitalism,
against which they contrast modern capitalism with its enormously increased amount of
state interventions, supposedly making more ideological legitimation necessary (Habermas,
op. cit., Ch. II.; OVe, Tauschverhaltnis und politische Steuerung. Zur Aktualitat des
Legitimationsproblems, in OVe, op. cit., pp. 2763). This view tends to veil the important role of ideology in the era of competitive capitalismthe era of human rights declarations, of the ascendance of bourgeois nationalism, and of still-strong established and
dissenting religionsand to veil as well the economic and Political mechanisms of crisis and revolution in the present period, a period which has witnessed the shattering of
the economic foundations of the British Empire and is witnessing the shaking of the
supremacy of the United States.
40. See the very important essay by Louis Althusser, Ideology and Ideological State
Apparatuses (Notes towards an Investigation), in his Lenin and Philosophy and Other Essays
(London: NLB, 1971), pp. 121273. For unconvincing reasons, however, Althusser talks
of ideological state apparatuses.
41. Marx, Das Kapital, III: 2, p. 140; Bottomore-Rubel, op. cit., p. 190.

Downloaded from crs.sagepub.com at UNIV MASSACHUSETTS AMHERST on September 15, 2012